Spaces:
Paused
Paused
File size: 12,487 Bytes
cd458ae | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 | # Copyright 2025 Alibaba Z-Image Team and The HuggingFace Team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from ...utils import logging
from ..modular_pipeline import AutoPipelineBlocks, SequentialPipelineBlocks
from ..modular_pipeline_utils import OutputParam
from .before_denoise import (
ZImageAdditionalInputsStep,
ZImagePrepareLatentsStep,
ZImagePrepareLatentswithImageStep,
ZImageSetTimestepsStep,
ZImageSetTimestepsWithStrengthStep,
ZImageTextInputStep,
)
from .decoders import ZImageVaeDecoderStep
from .denoise import (
ZImageDenoiseStep,
)
from .encoders import (
ZImageTextEncoderStep,
ZImageVaeImageEncoderStep,
)
logger = logging.get_logger(__name__) # pylint: disable=invalid-name
# ====================
# 1. DENOISE
# ====================
# text2image: inputs(text) -> set_timesteps -> prepare_latents -> denoise
# auto_docstring
class ZImageCoreDenoiseStep(SequentialPipelineBlocks):
"""
denoise block that takes encoded conditions and runs the denoising process.
Components:
transformer (`ZImageTransformer2DModel`) scheduler (`FlowMatchEulerDiscreteScheduler`) guider
(`ClassifierFreeGuidance`)
Inputs:
num_images_per_prompt (`None`, *optional*, defaults to 1):
TODO: Add description.
prompt_embeds (`list`):
Pre-generated text embeddings. Can be generated from text_encoder step.
negative_prompt_embeds (`list`, *optional*):
Pre-generated negative text embeddings. Can be generated from text_encoder step.
height (`int`, *optional*):
TODO: Add description.
width (`int`, *optional*):
TODO: Add description.
latents (`Tensor | NoneType`, *optional*):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
num_inference_steps (`None`, *optional*, defaults to 8):
TODO: Add description.
sigmas (`None`, *optional*):
TODO: Add description.
**denoiser_input_fields (`None`, *optional*):
The conditional model inputs for the denoiser: e.g. prompt_embeds, negative_prompt_embeds, etc.
Outputs:
latents (`Tensor`):
Denoised latents.
"""
block_classes = [
ZImageTextInputStep,
ZImagePrepareLatentsStep,
ZImageSetTimestepsStep,
ZImageDenoiseStep,
]
block_names = ["input", "prepare_latents", "set_timesteps", "denoise"]
@property
def description(self):
return "denoise block that takes encoded conditions and runs the denoising process."
@property
def outputs(self):
return [OutputParam.template("latents")]
# image2image: inputs(text + image_latents) -> prepare_latents -> set_timesteps -> set_timesteps_with_strength -> prepare_latents_with_image -> denoise
# auto_docstring
class ZImageImage2ImageCoreDenoiseStep(SequentialPipelineBlocks):
"""
denoise block that takes encoded text and image latent conditions and runs the denoising process.
Components:
transformer (`ZImageTransformer2DModel`) scheduler (`FlowMatchEulerDiscreteScheduler`) guider
(`ClassifierFreeGuidance`)
Inputs:
num_images_per_prompt (`None`, *optional*, defaults to 1):
TODO: Add description.
prompt_embeds (`list`):
Pre-generated text embeddings. Can be generated from text_encoder step.
negative_prompt_embeds (`list`, *optional*):
Pre-generated negative text embeddings. Can be generated from text_encoder step.
height (`None`, *optional*):
TODO: Add description.
width (`None`, *optional*):
TODO: Add description.
image_latents (`None`, *optional*):
TODO: Add description.
latents (`Tensor | NoneType`, *optional*):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
num_inference_steps (`None`, *optional*, defaults to 8):
TODO: Add description.
sigmas (`None`, *optional*):
TODO: Add description.
strength (`None`, *optional*, defaults to 0.6):
TODO: Add description.
**denoiser_input_fields (`None`, *optional*):
The conditional model inputs for the denoiser: e.g. prompt_embeds, negative_prompt_embeds, etc.
Outputs:
latents (`Tensor`):
Denoised latents.
"""
block_classes = [
ZImageTextInputStep,
ZImageAdditionalInputsStep(image_latent_inputs=["image_latents"]),
ZImagePrepareLatentsStep,
ZImageSetTimestepsStep,
ZImageSetTimestepsWithStrengthStep,
ZImagePrepareLatentswithImageStep,
ZImageDenoiseStep,
]
block_names = [
"input",
"additional_inputs",
"prepare_latents",
"set_timesteps",
"set_timesteps_with_strength",
"prepare_latents_with_image",
"denoise",
]
@property
def description(self):
return "denoise block that takes encoded text and image latent conditions and runs the denoising process."
@property
def outputs(self):
return [OutputParam.template("latents")]
# auto_docstring
class ZImageAutoDenoiseStep(AutoPipelineBlocks):
"""
Denoise step that iteratively denoise the latents. This is a auto pipeline block that works for text2image and
image2image tasks. - `ZImageCoreDenoiseStep` (text2image) for text2image tasks. -
`ZImageImage2ImageCoreDenoiseStep` (image2image) for image2image tasks. - if `image_latents` is provided,
`ZImageImage2ImageCoreDenoiseStep` will be used.
- if `image_latents` is not provided, `ZImageCoreDenoiseStep` will be used.
Components:
transformer (`ZImageTransformer2DModel`) scheduler (`FlowMatchEulerDiscreteScheduler`) guider
(`ClassifierFreeGuidance`)
Inputs:
num_images_per_prompt (`None`, *optional*, defaults to 1):
TODO: Add description.
prompt_embeds (`list`):
Pre-generated text embeddings. Can be generated from text_encoder step.
negative_prompt_embeds (`list`, *optional*):
Pre-generated negative text embeddings. Can be generated from text_encoder step.
height (`None`, *optional*):
TODO: Add description.
width (`None`, *optional*):
TODO: Add description.
image_latents (`None`, *optional*):
TODO: Add description.
latents (`Tensor | NoneType`):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
num_inference_steps (`None`):
TODO: Add description.
sigmas (`None`, *optional*):
TODO: Add description.
strength (`None`, *optional*, defaults to 0.6):
TODO: Add description.
**denoiser_input_fields (`None`, *optional*):
The conditional model inputs for the denoiser: e.g. prompt_embeds, negative_prompt_embeds, etc.
Outputs:
latents (`Tensor`):
Denoised latents.
"""
block_classes = [
ZImageImage2ImageCoreDenoiseStep,
ZImageCoreDenoiseStep,
]
block_names = ["image2image", "text2image"]
block_trigger_inputs = ["image_latents", None]
@property
def description(self) -> str:
return (
"Denoise step that iteratively denoise the latents. "
"This is a auto pipeline block that works for text2image and image2image tasks."
" - `ZImageCoreDenoiseStep` (text2image) for text2image tasks."
" - `ZImageImage2ImageCoreDenoiseStep` (image2image) for image2image tasks."
+ " - if `image_latents` is provided, `ZImageImage2ImageCoreDenoiseStep` will be used.\n"
+ " - if `image_latents` is not provided, `ZImageCoreDenoiseStep` will be used.\n"
)
# auto_docstring
class ZImageAutoVaeImageEncoderStep(AutoPipelineBlocks):
"""
Vae Image Encoder step that encode the image to generate the image latents
Components:
vae (`AutoencoderKL`) image_processor (`VaeImageProcessor`)
Inputs:
image (`Image`, *optional*):
TODO: Add description.
height (`None`, *optional*):
TODO: Add description.
width (`None`, *optional*):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
Outputs:
image_latents (`Tensor`):
video latent representation with the first frame image condition
"""
block_classes = [ZImageVaeImageEncoderStep]
block_names = ["vae_encoder"]
block_trigger_inputs = ["image"]
@property
def description(self) -> str:
return "Vae Image Encoder step that encode the image to generate the image latents"
+"This is an auto pipeline block that works for image2image tasks."
+" - `ZImageVaeImageEncoderStep` is used when `image` is provided."
+" - if `image` is not provided, step will be skipped."
# auto_docstring
class ZImageAutoBlocks(SequentialPipelineBlocks):
"""
Auto Modular pipeline for text-to-image and image-to-image using ZImage.
Supported workflows:
- `text2image`: requires `prompt`
- `image2image`: requires `image`, `prompt`
Components:
text_encoder (`Qwen3Model`) tokenizer (`Qwen2Tokenizer`) guider (`ClassifierFreeGuidance`) vae
(`AutoencoderKL`) image_processor (`VaeImageProcessor`) transformer (`ZImageTransformer2DModel`) scheduler
(`FlowMatchEulerDiscreteScheduler`)
Inputs:
prompt (`None`, *optional*):
TODO: Add description.
negative_prompt (`None`, *optional*):
TODO: Add description.
max_sequence_length (`None`, *optional*, defaults to 512):
TODO: Add description.
image (`Image`, *optional*):
TODO: Add description.
height (`None`, *optional*):
TODO: Add description.
width (`None`, *optional*):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
num_images_per_prompt (`None`, *optional*, defaults to 1):
TODO: Add description.
image_latents (`None`, *optional*):
TODO: Add description.
latents (`Tensor | NoneType`):
TODO: Add description.
num_inference_steps (`None`):
TODO: Add description.
sigmas (`None`, *optional*):
TODO: Add description.
strength (`None`, *optional*, defaults to 0.6):
TODO: Add description.
**denoiser_input_fields (`None`, *optional*):
The conditional model inputs for the denoiser: e.g. prompt_embeds, negative_prompt_embeds, etc.
output_type (`str`, *optional*, defaults to pil):
The type of the output images, can be 'pil', 'np', 'pt'
Outputs:
images (`list`):
Generated images.
"""
block_classes = [
ZImageTextEncoderStep,
ZImageAutoVaeImageEncoderStep,
ZImageAutoDenoiseStep,
ZImageVaeDecoderStep,
]
block_names = ["text_encoder", "vae_encoder", "denoise", "decode"]
_workflow_map = {
"text2image": {"prompt": True},
"image2image": {"image": True, "prompt": True},
}
@property
def description(self) -> str:
return "Auto Modular pipeline for text-to-image and image-to-image using ZImage."
@property
def outputs(self):
return [OutputParam.template("images")]
|